Reduced thalamic volume in high-functioning individuals with autism

Abstract

Background

In this study, specific consideration is given to a role for the thalamus in autism.

Methods

A volumetric analysis of the thalamus was conducted using magnetic resonance imaging, based on segmentation of continuous 1.2 mm³ coronal images. The sample consisted of 12 high-functioning individuals with autism, mean age of 21.0 years (SD = 10.4) and mean IQ of 106.4 (SD = 18.3). Normal control subjects were selected to match this group; the mean age was 18.1 years (SD = 6.3); mean IQ was 108.8 (SD = 15.6).

Results

Unadjusted mean thalamic volume was not significantly different; however, there were significant differences in the relationship between thalamic volume and total brain volume (TBV). The correlation was strong and positive in the control group but statistically nonsignificant in the autism group. Group differences were found when adjustments were made for TBV, achieved by grouping subjects’ measurements on this variable using a split median procedure. Mean thalamic volume was significantly reduced in the autism group relative to normal control subjects, specifically within the high TBV group.

Conclusions

The increase in thalamic volume with increase in TBV was not seen in autism, suggesting underdeveloped connections between cortical and subcortical regions and indicating a need to examine this structure further.

Introduction

Autism is widely acknowledged to be a neurodevelopmental disorder; however, the developmental events and neural mechanisms that underlie its expression are not well understood. Neuroimaging studies and models based on neuropathology, animal, and neuropsychologic data have focused on regions of the brain including the cerebellum Courchesne et al 1988, Courchesne et al 1994, Piven et al 1992, frontal cortex Ozonoff 1995, Zilbovicius et al 1995, hippocampus Bachevalier and Merjanian 1994, Bauman and Kemper 1985, Bauman and Kemper 1994, Piven et al 1998, and, more recently, the amygdala Abell et al 1999, Aylward et al 1999, Schultz et al 2000. There has also been increasing interest in a particular neural pathway, the cerebello-thalamo-cortical circuit, and its contribution in autism Carper and Courchesne 2000, Chugani et al 1997, Hardan et al 2001. Investigations of this pathway and its frontal and cerebellar connections have been conducted, but not of the thalamus specifically. Further consideration of a role for the thalamus in autism is merited. This structure may mediate specific functions relevant to autism and, of particular note, impact the organization of neural pathways and functional connectivity in the developing brain.

The thalamus forms a key component of neural systems, with reciprocal connections to virtually every major region of the brain Carpenter 1991, Kandel 1991, Katz and Shatz 1996. Each of the structures identified in the predominant neurologic theories of autism comprises a major neural pathway of which the thalamus is an integral part. Functionally, the thalamus has been implicated in domains such as attention LaBerge 1990, Paus et al 1997, Rafal and Posner 1987, memory (Bentivoglio et al 1997), language Crosson 1992, Nadeau and Crosson 1997, and emotional processing George et al 1995, Heilman 1997, Lane et al 1997, LeDoux 1986, LeDoux 1993, Reiman et al 1997, which are recognized areas of deficit in autism. The unusual sensory interests and stereotyped motor behaviors that are characteristic of autism also may be understood in terms of activity in the basal ganglia and thalamus. Above all, the thalamus is a relevant structure in autism given its proposed role for how information is received and processed in the brain Barth and MacDonald 1996, Castro-Alamancos and Connors 1996, Guillery 1995, Singer 1993, Steriade et al 1993 as well as its influence on cortical organization and differentiation during development Ghosh 1995, Molnár and Blakemore 1995, O’Leary 1989, O’Leary and Koester 1993. Whereas the thalamus has traditionally been referred to as the sensory gateway of the cortex, a current perspective of this structure suggests that it is involved in multiple processes that permit the transmission, tuning, and integrated processing of information in the brain. An examination of the thalamus may provide insight into autism as an information-processing disorder (e.g., Minshew et al 1997), particularly with regard to shaping the neural architecture of the brain in a way that leads to less functional connectivity and thus a less synchronized network.

In their seminal paper, Damasio and Maurer (1978) presented an extensive analysis of the patterns of abnormal behavior in autism and related these findings to forms of brain dysfunction in adults. They concluded that multiple brain regions are involved in autism, including the medial frontal lobes, medial temporal lobes, basal ganglia, and thalamus (specifically, the dorsomedial and anterior nuclear groups). In current functional imaging studies, preliminary data reveals less pronounced activation of the thalamus in individuals with autism in several domains. These include attention (Buchsbaum et al 1992), language (Mueller et al 1999), motor activation (Mueller et al 2001), and social processing (Baron-Cohen et al 1999) tasks. Disturbances in amygdalar–cortical systems have been speculated to be involved in autism, specifically with regard to transmitting information about emotional significance that guide social behaviors (Schultz et al 2000). As part of this pathway, the pulvinar nucleus of the thalamus may also participate in distinguishing relevant from irrelevant stimuli and signaling the emotional salience of visual stimuli Morris et al 1997, Robinson and Peterson 1992. Considered more broadly, the thalamus may have a role in directing attention to aspects of an experience, guided by particular motives and intentions Gaffan and Parker 2000, Johnson and Ojemann 2000, Nadeau and Crosson 1997, Newman 1995.

At present, there are no reports in the autism literature of a volumetric evaluation of the thalamus using magnetic resonance imaging (MRI). An older computerized tomography (CT) investigation with poor image resolution by current imaging standard reported a significant difference between left and right thalamic volume (L > R) in male adults with autism and normal control subjects but did not find a significant difference in thalamic size between the two groups (Creasey et al 1986). The groups were not matched on important variables, and levels of functioning were wide ranging in the autistic men. In the current investigation, we sought to conduct a volumetric evaluation of the thalamus in a well-characterized and well-matched sample, with enhanced spatial resolution and carefully defined structural boundaries to determine if its gross anatomy differed in comparison with matched control subjects. In addition, we investigated the relationship between thalamus and total brain volume in individuals with autism and normal controls.